Hydroclimatological Teleconnections from Land-Cover Change in Tropical Regions

Past studies have indicated that deforestation of the Amazon basin would result in an important rainfall decrease in that region, but that this process had no significant impact on the global temperature or precipitation and had only local implications. Here we show that deforestation of tropical regions significantly affects precipitation at mid and high latitudes through hydrometeorological teleconnections. In particular, we find that the deforestation of Amazonia and Central Africa severely reduces rainfall in the lower US Midwest during the spring and summer seasons and in the upper US Midwest during the winter and spring, respectively, when water is crucial for agricultural productivity in these regions. Deforestation of South-East Asia affects most significantly China and the Balkan Peninsula. On the other hand, the elimination of any of these tropical forests considerably enhances summer rainfall in the southern tip of the Arabian Peninsula. The combined effect of deforestation of these three tropical regions causes a significant decrease in winter precipitation in California and seems to generate a cumulative enhancement of precipitation during the summer in the southern tip of the Arabian Peninsula. We also use regional and global climate models in conjunction with socio-economic scenarios of land use / land-cover change in the Amazon basin, to estimate potential changes in the water cycle inside and outside of the basin. Four different experiments were produced with the NASA-GISS Global Climate Model (GCM): (1) a "current land cover" ensemble, which also serves as the "control" ensemble; (2) a "scenario for 2030" ensemble; (3) a "scenario for 2050" ensemble; and (4) a "total deforestation" ensemble that simulates the land cover in the Amazon basin after all the tropical forest has been eliminated. In addition, The Regional Atmospheric Modeling System (RAMS) is used at a high resolution (20-km grid size) and very-high resolution (1-km grid size) over the Amazon Basin and using the same four land-cover scenarios with the NCEP reanalysis for four different years (wet - 1997, dry - 1998, and two "normal" years – 1999 and 2000 that have similar domain-average precipitation but different spatial distributions) forcing its lateral boundaries. Thus, the combined impacts of deforestation and El Nino and La Nina years are also explored as part of this numerical experiment. The combination of these different simulations reveals significant impact of deforestation on the regional and global hydroclimate through land-cover change teleconnections.